Slashdot Mirror

← Back to Stories (view on slashdot.org)

Microsoft Sued Over TCP/IP offload technology

Posted by Hemos on from the the-legal-battles-begin dept.

soldack writes "Microsoft has been working on a technology for full offload of TCP/IP processing in windows to a smart NIC. It was due to arrive in an update to Windows Server 2003 but never made it. Now know why: it appears they have been sued by Alacritech for patent violation and Alacritech has won an injunction. See this article on Microsoft Watch for a story on it and here for Alacritech's view on it. It is pretty interesting to see a little company trying to take on Microsoft for seemingly ripping them off and getting this far. It probably helps that they were founded by Larry Boucher, who "led the engineering team that developed the SCSI interface at Shugart Associates" and also founded Adaptec. See Alacritech's site for more. Lots of TOE/RNIC companies were effected by this since they were depending on Microsoft's software and do not have their own solution. This technology is becoming more important as the industry moves to multiple 1 gigabit interfaces and single or multiple 10 gigabit interfaces. It may be critical for technologies like iWarp (RDMA over TCP/IP) and iSCSI (SCSI over TCP/IP) to perform well."

13 of 19 comments (clear)

  1. Greenmail by redelm · · Score: 1
    This sounds like just another step in royalty / licencing negotiations.

    1. Re:Greenmail by Saeed+al-Sahaf · · Score: 1, Insightful

      Exactly. And of course the SlashSnot crowd is totally hypocritical on this because they are too blinded by anger at the big bad Microsoft to realize that this is just another obnoxious software patent. The "my enemy's enemies are my friend" is typical selfish crap that devalues clear thinking.

      --
      "Who are in control, they are not in control of anything - they don't even control themselves!" - Glen Beck
    2. Re:Greenmail by truthsearch · · Score: 1

      Generally I'd agree with you. And I'm entirely against software patents. But consider Microsoft's actions while software patents are legal:

      "Alacritech claims that it discussed its technology with Microsoft in 1998 and that Microsoft subsequently cut off communication with the company. In May 2003, Microsoft demonstrated a technology it called "Chimney" that Alacritech said was similar to its own intellectual property. Alacritech offered Microsoft a license, but Microsoft rejected Alacritech's terms, and in August 2004, Alacritech filed suit claiming patent infringement."

      Microsoft's used this tactic before and lost in court. They meet, often make an agreement, then close their doors and duplicate the technology from another company. If it weren't over a software patent wouldn't you still want that practice to be stopped (assuming they signed an NDA or other contract)?

      --
      Developers: We can use your help.
    3. Re:Greenmail by RogueLeaderX · · Score: 1

      Let's consider a hypopethetical world where software patents are illegal and Microsoft tried this. Alacritech sues Microsoft for breach of contract and wins. Some result, no messy patent system. Not a critique of your statement, a critique of the patent system.

    4. Re:Greenmail by drsmithy · · Score: 1
      Let's consider a hypopethetical world where software patents are illegal and Microsoft tried this. Alacritech sues Microsoft for breach of contract and wins.

      In your hypothetical world there wouldn't be anything they could sue them about - Microsoft going off and reimplementing someone else's software on their own would be perfectly legal.

  2. Been done. by bluGill · · Score: 3, Interesting

    Not with TCP/IP, but I've seen (not worked with) boards to offload network processing from the host machine. They ran in mainframes or VAXen, and other such large machines about 20 years ago. Never caught on though because you still need a reliable protocol to talk to the offload processor. (I was only aware that a box up on the top shelf contained such things, I'm not sure how they work, but could be prior art)

    1. Re:Been done. by BrookHarty · · Score: 1

      This is more than "A network card with a CPU", its also the technology "SLIC" to go along with the ASIC card. SLIC was copied by Microsoft after meeting with Alacritech, and then Microsoft saying they invented it, and broke all communications.

      Now, Microsoft might say that the technology already existed, but its too specialized, with parts on the CPU and parts on the Network card. I suspect Microsoft will just license the technology and end suit but not after trying years of legal bills.

      http://www.alacritech.com/html/tech_review.html has details on the technology.

    2. Re:Been done. by superpulpsicle · · Score: 1

      Well about the iSCSI part, I think it almost doesn't matter. Fibre-channel is clearly the dominate protocol in the storage industry right now.

  3. HPUX by Asgard · · Score: 2, Informative

    I seem to recall hearing that the old HPUX machines had a sort of smart-NIC; the machine could be completely crashed and yet ping still worked.

  4. Different from 3COM? by snookerdoodle · · Score: 1

    'Just an honest question (ok, 2 questions): don't some of 3COM's NIC's do something similar? Is the difference just a matter of the degree of offloading?

    Mark

    1. Re:Different from 3COM? by soldack · · Score: 2, Informative

      Windows (and Linux for that matter) support offloading checksum calculations. Windows also support large sends where segmentation is offloaded. Finally windows supports ipsec offload.
      All of the top NICs (3com, broadcom, intel) support this. Microft was working on offloading all of TCP/IP processing. In theory, the NICs have ASICs and/or network processors that can handle high speed (think > 1 gigabit) TCP/IP processing and the OS just gives it data to send and a place to receive.

      Depending on what you are doing it can make a big difference. Segmentation/reassembly can cost a lot of CPU time. Its effect is less if both sides of the connection (and the switches in between) support jumbo frames (9KB frames instead of 1500 byte frames). If done properly TCP/IP offload can even lead to avoiding the user/kernel switch and the copying of network frames that waste memory bandwidth. Many folks have noted that checksum offload doesn't help much when the memory is still being copied. Also, in theory an ASIC could handle TCP/IP with lower latency and higher throughput than most server CPUs. Other network adapters targeted at high performance computing like InfiniBand have their form of TCP/IP embedded in the ASICs of the controllers and show > 10 gigabit bidirectional throughput and sub 5 microsecond end-to-end latency in PCI Express x8 slots.

      --
      -- soldack
  5. Dupe by jleq · · Score: 1

    Duplicate post, see http://yro.slashdot.org/article.pl?sid=05/04/13/21 59255

  6. What the PDP's used in 1969 by ScrappyLaptop · · Score: 1
    Honeywell DDP-516's a.k.a. IMP's. An entire freestanding computer just to offload the network traffic! So robust that you could hook just the IMPs up to the 'network' and have them communicate amongst themselves -WITHOUT THEIR HOST COMPUTERS!

    A little perspective: A $10 NIC is 'smart' enough with the system turned off to monitor traffic for smart packets meant for its MAC...